A Robust and Versatile Photoinduced Living Polymerization of

Apr 1, 2014 - School of Chemical Engineering, UNSW Australia, Sydney, NSW 2052, Australia. •S Supporting ... applications in high-tech industry. How...
0 downloads 0 Views 680KB Size
Download PDF
Article pubs.acs.org/JACS

A Robust and Versatile Photoinduced Living Polymerization of Conjugated and Unconjugated Monomers and Its Oxygen Tolerance Jiangtao Xu,†,‡ Kenward Jung,† Amir Atme,‡ Sivaprakash Shanmugam,† and Cyrille Boyer*,†,‡ †

Centre for Advanced Macromolecular Design (CAMD), School of Chemical Engineering, and ‡Australian Centre for NanoMedicine, School of Chemical Engineering, UNSW Australia, Sydney, NSW 2052, Australia S Supporting Information *

ABSTRACT: Controlled/living radical polymerization techniques have transformed polymer chemistry in the last few decades, affording the production of polymers with precise control over both molecular weights and architectures. It is now possible to synthesize almost an infinite variety of macromolecules using nonspecialized equipment, finding applications in high-tech industry. However, they have several shortcomings. Until recently, living radical polymerizations could not be controlled by an external stimulus, such as visible light, pH, mechanical, chemical, etc. Moreover, they are usually sensitive to trace amounts of oxygen in the system. In this Article, we report a photoinduced living polymerization technique, which is able to polymerize a large range of monomers, including conjugated and unconjugated monomers, using ultralow concentrations of an iridium-based photoredox catalyst (typically 1 ppm to monomers) and a low energy visible LED as the light source (1−4.8 W, λmax = 435 nm). The synthesis of homopolymers with molecular weights ranging from 1000 to 2 000 000 g/mol was successfully achieved with narrow molecular weight distributions (Mw/Mn < 1.3). In addition, chain extensions of poly(methacrylate)s, poly(styrene), poly(N-vinyl pyrrolidinone), poly(vinyl ester)s, and poly(acrylate)s were performed to prepare diblock copolymers. The reusability of the catalyst was demonstrated by the synthesis of a decablock polymer by multiple chain extensions. Most importantly, this process was employed to prepare well-defined polymers and multiblock copolymers in the presence of air.

■

workers3 developed an analogue of atom transfer radical polymerization (ATRP) controllable by light in 2012. A versatile photocontrolled polymerization technique able to polymerize a broader range of monomers, including both conjugated monomers ((meth)acrylate, (meth)acrylamide, and styrene) and unconjugated monomers (vinyl esters, N-vinyl pyrrolidinone, and dimethyl vinylphosphonate), preferably under low energy visible light (low wattage), would be desirable. We envisioned a new photoinduced living polymerization technique that would be able to polymerize a large range of monomers, even in the presence of oxygen. In this technique, a photoredox catalyst (in our case, fac-[Ir(ppy)3], Ir(III), Supporting Information, Scheme S1) was employed to generate an excited species (Ir(III)*) under irradiation, which was then able to reduce thiocarbonylthio compounds via photoinduced electron transfer (PET),37 resulting in the production of radicals (Pn•) and Ir(IV) species (Scheme 1). In contrast to the conventional RAFT mechanism, the thiocarbonylthio compound in this technique acted as an initiator in addition to being a chain transfer agent. The radical (P•), generated by the

INTRODUCTION The development of controlled polymerization techniques triggerable by stimulus, including photochemical, thermal, and electrochemical stimuli, has grown in interest due to their unique abilities to be switched between “ON” and “OFF” states.1−16 Since the seminal work of Otsu and co-workers,17 radical polymerization triggered by light has attracted a great deal of attention due to the ease in control of polymerizations in both space and time.10,18−20 It has been demonstrated that this technique can be used to prepare various architectures, such as block, graft, and star polymers. Recently, with the emergence of reversible addition−fragmentation chain transfer (RAFT) polymerization,21−24 thiocarbonylthio compounds were successfully employed for photocontrolled radical polymerization in the presence or absence of photoinitiators.10,18,19,25,26 Despite some success in employing these techniques, they were not without limitations, most notably, the loss of end-group fidelity due to the photolysis of the RAFT end-group under UV light.27−29 Inspired by the early works of Stephenson,30−34 Yoon,35,36 and Macmillan37,38 on the atom transfer radical addition (ATRA) of halide compounds onto olefins catalyzed by a photoredox catalyst and by the recent works of Lalevée on free radical polymerization using photocatalysts, Hawker and co© 2014 American Chemical Society

Received: February 18, 2014 Published: April 1, 2014 5508

dx.doi.org/10.1021/ja501745g | J. Am. Chem. Soc. 2014, 136, 5508−5519

Journal of the American Chemical Society

Article

Scheme 1. Proposed Mechanism of a Photoinduced Living Polymerization Using fac-[Ir(ppy)3] as Photoredox Catalyst

Scheme 2.

a

a

List of monomers investigated in this study: (a) methyl methacrylate (MMA), (b) methyl acrylate (MA), (c) tert-butyl acrylate (tBuA), (d) styrene (St), (e) N,N-dimethylacrylamide (DMA), (f) N-(2-hydroxypropyl) methacrylamide (HPMA), (g) N-isopropylacrylamide (NIPAAm), (h) oligoethylene glycol methyl ether methacrylate (OEGMA), (i) oligoethylene glycol methyl ether acrylate (OEGA), (j) isoprene, (k) vinyl acetate, (l) vinyl pivalate (VP), (m) N-vinyl pyrolidinone (NVP), (n) dimethyl vinylphosphonate (DVP).

reduction of thiocarbonylthio compound via PET process, could initiate polymerization of the monomers. The radical (P•) may then participate in the RAFT process or be deactivated by Ir(IV) to regenerate the initial Ir(III), which would restart the catalytic cycle. Thus, this process involves an electron transfer process between the photocatalyst and the thiocarbonylthio compound and a cooperative RAFT process. This technique presents several further merits in comparison to the conventional RAFT mechanism. The polymerization reactions can be performed at room temperature using low energy blue light (1−4.8 W, λmax = 435 nm) in tandem with catalyst doses in the ppm range. More importantly, the elimination of exogenous radical initiators limits the formation of dead polymers. We employed this photoinduced living polymerization technique for the controlled polymerization of a range of different monomers, including methacrylates, acrylates, styrene, acrylamides, methacrylamides, vinyl esters, vinyl phosphonate, and N-vinyl pyrrolidinone monomers (Scheme 2). The synthesis of multiblock polymers via successive chain extensions was also achieved to yield complex block polymers. Finally, we exploited the strong reductive properties of fac[Ir(ppy)3] to activate the polymerization without prior deoxygenation processes, which resulted in the synthesis of well-defined polymers with narrow molecular weight distributions.

were performed. fac-[Ir(ppy)3] catalyst (Ir(III)) is a fluorescent compound, with an excitation and emission wavelength at 376 and 524 nm, respectively, in DMSO (Supporting Information, Figure S3). In these quenching experiments, the concentration of the thiocarbonylthio compound in a solution of fac[Ir(ppy)3] was gradually increased, and the intensity of the fluorescence emission peak at 524 nm was measured. We observed a decrease in the emission intensity (I) after addition of the thiocarbonylthio compound (Figure 1A), which confirmed energy transfer between fac-[Ir(ppy)3] and the thiocarbonylthio compound. Plotting the ratio I0/I versus the quencher concentration showed a nonlinear relationship, indicative of both dynamic and static quenching behaviors. In the case of dynamic quenching (also called collisional quenching), the excited state of fac-[Ir(ppy)3] transfers the energy to the thiocarbonylthio compound, whereas static quenching results in the formation of a complex (Figure 1B). 2. Polymerization of Methyl Methacrylate by Photoinduced Living Polymerization. After proving that energy transfer occurred between the thiocarbonylthio compounds and excited Ir(III)*, we decided to test our process for the polymerization of methyl methacrylate (MMA) using a dithiobenzoate (CPADB, Scheme 1) as initiator and chain transfer agent, fac-[Ir(ppy)3] as catalyst, and a 4.8 W blue LED light source in N,N′-dimethylformamide (DMF) (Table 1, no. 1). Poly(methyl methacrylate) (PMMA) polymers were obtained with relatively good control of the molecular weights and narrow molecular weight distributions (MWDs). These results motivated us to reduce the concentration of catalyst to 1 ppm (Table 1, no. 2). After 67 h, 73% monomer conversion was determined by nuclear magnetic resonance (NMR), demonstrating that the successful polymerization could be performed at very low concentrations of catalyst. Although decreasing the catalyst concentration also resulted in slightly slower polymerization kinetics, such ultralow concentration of catalyst is potentially desirable for industrial applications. Inspired by the initial work of Stephenson and co-workers34 on the efficiency of the catalyst for organic reactions, we

■

RESULTS AND DISCUSSION 1. Investigation of Photoinduced Living Polymerization Mechanism Using Fluorescence Spectroscopy and Cyclic Voltammetry. The proposed mechanism for photoinduced living polymerization described in Scheme 1 is based on a photoredox reaction between a thiocarbonylthio compound and a photoredox catalyst. The redox potential of Ir(IV)/Ir(III)* is −1.72 V (potential versus the saturated calomel electrode (SCE) in acetonitrile),34,39 whereas the redox potentials of thiocarbonylthio compounds are usually higher,40,41 −0.4, −0.6, and −0.9 V (potential versus SCE) for CPADB, BTPA, and BSTP, respectively (Supporting Information, Figure S2). As the redox potentials are all greater than −1.72 V, we expected that Ir(III)* will be able to reduce the thiocarbonylthio compounds via the photoelectron transfer (PET) process. To prove the electron transfer between fac[Ir(ppy)3] and the thiocarbonylthio compound did occur, fluorescence quenching (or Stern−Volmer quenching) studies 5509

dx.doi.org/10.1021/ja501745g | J. Am. Chem. Soc. 2014, 136, 5508−5519

Journal of the American Chemical Society

Article

Figure 1. Fluorescence quenching studies of a 7.64 μM solution of fac-[Ir(ppy)3] in DMSO with varying concentrations of thiocarbonylthio compounds (CPADB, BTPA, BSTP, and xanthate). (A) Fluorescent emission intensity versus different concentrations of CPADB; (B) plots of the ratio Io/I versus quencher concentration. Io and I correspond to the emission intensity in the absence and presence of quencher, respectively.

Table 1. Examples of PMMA Synthesized in This Study no. 1 2 3 4 5 6

exp. cond.a [M]:[CPADB]:[Ir] −4

200:1:10 × 10 200:1:2 × 10−4 200:1:4 × 10−4 200:1:2 × 10−4 4000:1:40 × 10−4 200:1:0

solvent

[Ir]/[M] (ppm)

time (h)

αb (%)

Mn,thc (g/mol)

Mn,GPCd (g/mol)

Mw/Mnd

DMF DMF DMSO DMSO DMSO DMSO

5 1 2 1 1 0

23 67 24 36 24 48

69 73 71 85 88 0

14 120 14 520 14 620 17 260 352 200 0

14 300 14 700 (15 100)e 14 100 17 000 335 000

1.18 1.09 1.12 1.09 1.23

a The reactions were performed at room temperature under 4.8 W blue LED light (λmax = 435 nm). bMonomer conversion determined by 1H NMR spectroscopy was calculated by the following equation: α = (1 − [(I5.5−6.0ppm/2)/(I3.5ppm/3)]) × 100. cTheoretical molecular weight calculated using the following equation: Mn,th = [MMA]0/[CPADB]0 × MWMMA × α + MWCPADB, where [MMA]0, [CPADB]0, MWMMA, α, and MWCPADB correspond to MMA and CPADB concentration, molar mass of MMA, monomer conversion, and molar mass of CPADB. dMolecular weight and polydispersity determined by GPC analysis (DMAc used as eluent). eMolecular weight determined by 1H NMR using Mn,NMR = (I3.5ppm/3)/ (I7.2−7.8ppm/5) × MWMMA + MWCPADB, where I3.5ppm and I7.2−7.8ppm correspond to integrals of signal at δ 3.5 ppm and δ 7.2−7.8 ppm attributed to OCH3 of MMA and phenyl group of CPADB, respectively.

Figure 2. Photopolymerization of MMA using CPADB and fac-[Ir(ppy)3] as photoredox catalyst in the presence (“ON”) or in the absence (“OFF”) of light: (a) conversion versus time; (b) ln([M]0/[M]t) versus time of exposure; (c) Mn (●) and Mw/Mn (○) versus conversion; and (d) GPC traces at different times of exposure. Room temperature; 4.8 W blue LED light; [MMA]:[CPADB]:[catalyst] = 200:1:10 × 10−4.

synthesis,6,8,32,42 we observed that DMSO allowed for significantly faster polymerization kinetics and lower Mw/Mn values, suggesting higher catalytic efficiency as compared to other solvents (Table 1, nos. 3−5). Different Mn’s of PMMA were prepared in DMSO ranging from 2500 to 350 000 g/mol with narrow MWDs, Mw/Mn < 1.30, validating the versatility of

initiated the exploration of other solvents, including dimethyl sulfoxide (DMSO), acetonitrile, toluene, and methanol (Supporting Information, Tables S1 and S2). Successful polymerizations were obtained for all of these solvents, which demonstrated the versatility of this polymerization technique. Consistent with previous observations in organic and polymer 5510

dx.doi.org/10.1021/ja501745g | J. Am. Chem. Soc. 2014, 136, 5508−5519

Journal of the American Chemical Society

Article

Figure 3. Photopolymerization of MMA using (A) varied concentrations of photoredox catalyst and (B) different light intensities of blue LED (λmax = 435 nm) in the presence of CPADB at room temperature, using a molar ratio of [MMA]:[CPADB] = 200:1 in DMSO.

Table 2. Examples of Polymers Synthesized Using Conjugated Monomers in This Study no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

exp. cond.a [M]:[thiocar]:[Ir] −4

200:1:10 × 10 400:1:20 × 10−4 200:1:2 × 10−4 200:1:10 × 10−4 200:1:2 × 10−4 200:1:4 × 10−5 200:1:2 × 10−5 200:1:2 × 10−4 40:1:2 × 10−4 200:1:10 × 10−4 200:1:20 × 10−4 200:1:20 × 10−3 200:1:2 × 10−4 200:1:2 × 10−4 200:1:10 × 10−4

monomer HPMA HPMA HPMA MA MA MA MA MA OEGA tBA St St DMA NIPAAm isoprene

thiocar CPADB CPADB CPADB BTPA BTPA BTPA BTPA BSTP BTPA BTPA BTPA BTPA BTPA BTPA BTPA

[Ir]/[M] (ppm) 5 5 1 5 1 0.2 0.1 1 5 5 10 100 1 1 5

time (h) 24 24 24 2 2 10 10 2 8 2 48 24 3 3 12

αb (%) 70 72 21 99 93 96 83 88 81 88 50 89f 93 95 23

Mn,thc (g/mol) 20 300 41 000 6280 17 800 16 170 16 770 14 530 14 640 14 630 25 480 9630 18 810f 23 740 21 700 6200

Mn,GPCd (g/mol) e

58 600 (24 210) 49 200 12 100 (6490)e 17 500 15 500 17 100 15 300 15 430 15 800 22 200 9100 18 700f 18 430 21 400 3200

Mw/Mnd 1.16 1.08 1.09 1.05 1.08 1.12 1.19 1.13 1.11 1.09 1.13 1.21f 1.09 1.08 1.35

a

The reactions were performed at room temperature under 4.8 W blue LED light (λmax = 435 nm). bMonomer conversion for MA determined by 1H NMR spectroscopy was calculated by the following equation: α = (1 − [(I5.5−6.0ppm/2)/(I3.5ppm/3)]) × 100. cTheoretical molecular weight calculated using the following equation: Mn,th = [M]0/[thiocar]0 × MWM × α + MWthiocar, where [M]0, [thiocar]0, MWM, α, and MWthiocar correspond to monomer and thiocarbonylthio compound concentration, molar mass of monomer, monomer conversion, and molar mass of thiocarbonylthio compound. dMolecular weight and polydispersity determined by GPC analysis. eMolecular weight determined by 1H NMR using Mn,NMR = (I3.8ppm/ 1)/(I7.8ppm/2) × MWHPMA + MWCPADB, where I3.8ppm and I7.8ppm correspond to integrals of signal at δ 3.8 ppm and δ 7.8 ppm attributed to CH of HPMA and phenyl group (Z-group) of CPADB. fPSt phase separates at high monomer conversion in DMSO due its poor solubility in this solvent, resulting in a slight increase of Mw/Mn.

Table 3. Examples of Polymers Synthesized Using Unconjugated Monomers in This Study no. 1 2 3 4 5 6 7 8 9 10 11 12

exp. cond.a [M]:[thiocar]:[Ir] −4

200:1:40 × 10 200:1:40 × 10−4 200:1:10 × 10−4 200:1:10 × 10−4 200:1:2 × 10−4 1000:1:50 × 10−4 200:1:10 × 10−4 200:1:10 × 10−4 100:1:10 × 10−4 100:1:10 × 10−4 170:1:17 × 10−4 170:1:17 × 10−4

monomer

thiocar

[Ir]/[M] (ppm)

time (h)

αb (%)

Mn,thc (g/mol)

Mn,GPCd (g/mol)

Mw/Mn

VAc VAc VAc VAc VAc VAc VP VP DVP DVP NVP NVP

BTPA xanthate xanthate xanthate xanthate xanthate xanthate xanthate xanthate xanthate xanthate xanthate

20 20 5 5 1 5 5 5 10 10 10 10

24 22 2 24 20 22 3 24 6 14 6 14

0 76 16 81 41 NDe 22 80 22 41 40 65

13 300 3700 14 000 7200 NDe 4500 20 800 3100 5800 6900 12 500

18 200 5300 18 300 11 900 56 000 3800 22 000 3500 6700 7200 13 200

1.20 1.09 1.20 1.18 1.38 1.18 1.38 1.27 1.17 1.23 1.10

a

The reactions were performed at room temperature under 4.8 W blue LED light (λmax = 435 nm). bMonomer conversion determined by 1H NMR spectroscopy. cTheoretical molecular weight calculated using the following equation: Mn,th = [M]0/[xanthate]0 × MWM × α + MWxanthate, where [M]0, [xanthate]0, MWM, α, and MWxanthate correspond to M and xanthate concentration, molar mass of M, monomer conversion, and molar mass of xanthate. dMolecular weight and polydispersity determined by GPC analysis. eND: not determined.

this technique (Supporting Information, Table S1). To confirm that the activation and deactivation is induced only by the catalyst under light, control experiments in the absence of

catalyst (Table 1, no. 6) or light (data not shown) were conducted. Polymerization was not detected in either case. 5511

dx.doi.org/10.1021/ja501745g | J. Am. Chem. Soc. 2014, 136, 5508−5519

Journal of the American Chemical Society

Article

rates with little effect on the polydispersities and the control of molecular weights. In contrast to the radical initiators used in conventional RAFT and ATRP polymerization, the amount of catalyst employed in photoinduced living polymerization is much lower with a typical ratio of 1 ppm catalyst relative to monomer. For instance, ATRP48−50 or single electron transferliving radical polymerization (SET-LRP)12,51,52 typically employs a ratio of [monomer]:[copper] equal to 100:148 (except in the case of activator regenerated by electron transfer (ARGET-ATRP) ATRP, where a lower ratio can be employed under specific conditions),53−55 while in the RAFT process a ratio of [thiocarbonylthio]:[initiator] equal to 100:10 is commonly employed.22−24,50,56−60 To further explore the efficiency of the catalyst, the light intensity was varied from 1.0 to 4.8 W. Although lower light intensities (1.0 and 2.4 W) resulted in both a slight increase of the inhibition period and a slight decrease in the polymerization rate (Figure 3B), the polymerizations still exhibited good control of the molecular weights and polydispersities ( 100 000 g/mol). 5513

dx.doi.org/10.1021/ja501745g | J. Am. Chem. Soc. 2014, 136, 5508−5519

Journal of the American Chemical Society

Article

Figure 6. (A) Comparison of molecular weight distributions recorded using a RI and UV (λ = 305 nm) detector and (B) 1H NMR spectra for purified PMMA and PMA polymer synthesized by photoinduced living polymerization in the presence of air using BTPA and 4.8 W blue LED lamp (λmax = 435 nm) as light source (Mn,NMR,PMMA = 8010 g/mol, Mn,GPC,PMMA = 8200 g/mol, monomer conversion 41%; Mn,NMR,PMA = 4620 g/mol, Mn,GPC,PMA = 4700 g/mol, monomer conversion 29%).

All three techniques displayed high end-group fidelity (∼100%). The controlled/living character was further demonstrated by monitoring monomer conversion and molecular weight versus exposure time for both MA and St (Supporting Information, Figures S14 and S18). 1H NMR, GPC using RI and UV detector, and UV−vis were also employed to confirm the presence of trithiocarbonate end-group for a PSt polymer (Supporting Information, Figures S19,20). In addition, good control of the molecular weight and a narrow MWD was also noted for BSTP (Table 2, no. 8) with a polymerization rate comparable to that of BTPA, suggesting that this polymerization has the potential to utilize a broad variety of trithiocarbonate compounds. Other acrylic monomers, such as OEGA and tBA, were also successfully polymerized using this technique (Table 2, nos. 9,10). Subsequently, styrene was investigated using different concentrations of catalyst. The polymerization of styrene required a high amount of photoredox catalyst to reach high conversion. We attributed the slow

polymerization of MA can be carried out using an ultralow concentration of catalyst. As expected, at lower catalyst concentration (0.1 ppm), the polymerization required longer reaction time to reach high monomer conversion. Additionally, a lower catalyst concentration resulted in a slight increase of the PDI. Subsequently, the concentration of BTPA was varied to prepare PMA polymers with different molecular weights ranging from 2000 to 2 000 000 g/mol (Supporting Information, Table S4). To our knowledge, the synthesis of high molecular weight polyacrylate polymers using controlled/living radical polymerization techniques, including RAFT, ATRP, and NMP, is challenging and has only been successfully reported by a limited number of specific techniques.52,64−66 To confirm the presence of the trithiocarbonate end-group, PMA with molecular weight of 8000 g/mol was purified and analyzed by 1H NMR (Supporting Information, Figure S15) and UV−vis spectroscopy (Supporting Information, Figures S16 and S17). 5514

dx.doi.org/10.1021/ja501745g | J. Am. Chem. Soc. 2014, 136, 5508−5519

Journal of the American Chemical Society

Article

Figure 7. Kinetics plots for photoinduced living polymerizations of MMA (A and B) and MA (C and D) in the presence of oxygen (red ●) and absence of oxygen (■) in DMSO. (A) ln([M]0/[M]t) versus exposure time for MMA; (B) Mn versus conversion (top) and Mw/Mn versus conversion (bottom) for MMA; (C) ln([M]0/[M]t) versus exposure time for MA; and (D) Mn versus conversion (top) and Mw/Mn versus conversion (bottom) for MA. Note: [M]0 and [M]t correspond to the concentrations of monomers at time zero and t, respectively. (B and D) Straight lines correspond to the theoretical values.

polymerization of styrene to its low propagation rate constant (kp) at room temperature. However, at high conversion, we observed that PSt precipitated in DMSO due to its poor solubility in this solvent. Nevertheless, good control was still achieved. In contrast, acrylamide, including DMA and NIPAAm, was rapidly polymerized in 3 h (close to full monomer conversion) using a very low amount of catalyst (1 ppm relative to monomer). As indicated by GPC and NMR, the molecular weights were well controlled with a narrow MWD (Mw/Mn < 1.10). Finally, isoprene (Table 2, no. 15) was investigated using BTPA as initiator and chain transfer agent in the presence of photoredox catalyst (5 ppm). After 12 h, we observed a conversion of 23% determined by NMR, while GPC revealed the synthesis of polyisoprene with a molecular weight of 3200 g/mol (and Mw/Mn of 1.35), which demonstrates that this process is very efficient for the polymerization of conjugated monomers. 4. Polymerization of Unconjugated Monomers, Including Vinyl Acetate (VAc), Vinyl Pivalate (VP), NVinyl Pyrrolidinone (NVP), and Dimethyl Vinylphosphonate (DVP), by Photoinduced Living Polymerization. We decided to investigate unconjugated monomers using our process. In this work, we selected four different model monomers, including vinyl acetate (VAc), vinyl pivalate (VP), N-vinylpyrrolidinone (NVP), and dimethyl vinylphosphonate (DVP). All of these monomers are widely employed in industry due to their interesting properties. For instance, PVAc is the precursor of polyvinyl alcohol (used in coatings and also a biocompatible polymer), and PNVP is used in the synthesis of inks, coatings, and adhesives. First, vinyl acetate (VAc) was investigated using BTPA or xanthate in the presence of various photoredox catalyst concentrations. Initial polymerizations using BTPA as initiator and chain transfer agent were

unsuccessful (Table 3, no. 1). However, successful polymerizations were obtained with xanthate (Table 3, nos. 2−6). The experimental molecular weights determined by GPC were greater than the theoretical values, which was attributed to the difference in hydrodynamic volume between PVAc and the PSt standard. NMR was invoked to calculate the molecular weight. Mn,NMR was in good agreement with the theoretical values. Interestingly, the amount of photoredox catalyst does not affect the molecular weight distribution, as all of the polymerizations displayed a PDI lower than 1.20. After these initial successful results, VAc kinetics was investigated using [fac-[Ir(ppy)3]]/ [monomer] of 5 ppm. A linear evolution of ln([M]0/[Mt]) and molecular weight versus exposure time demonstrates the living character of this polymerization (Figure 4). To demonstrate the presence of xanthate end-group, PVAc (Table 3, nos. 2 and 3) was analyzed by NMR (Supporting Information, Figure S21) and GPC equipped with RI and UV detector (Supporting Information, Figure S22). Other monomers, including vinyl pivalate (VP), N-vinyl pyrrolidinone (NVP), and dimethyl vinylphosphonate (DVP), were also tested using a catalyst concentration of 10 ppm (relative to monomer). These monomers revealed the synthesis of polymers with a narrow MWD and good control of the molecular weight (Table 3, nos. 7−12). These results demonstrate that this polymerization technique can control a diverse range of unconjugated monomers. 5. Synthesis of Diblock Copolymers Using Different Monomer Families. To investigate the versatility of this approach, we decided to prepare block polymers comprised of different monomer families. Six different macroinitiators, that is, PMMA, PHPMA, PSt, PMA, PDMA, and PNVP, were prepared by photoinduced living polymerization, and subsequently purified by precipitation (Table 4). First, the PMMA 5515

dx.doi.org/10.1021/ja501745g | J. Am. Chem. Soc. 2014, 136, 5508−5519

Journal of the American Chemical Society

Article

4, no. 9). Such results have been previously reported in the literature for RAFT and ATRP process and are attributed to the difference in reactivity between the end-groups (PMMA-RAFT and PSt-RAFT). Successful chain extension of PSt with MA was confirmed by GPC with a narrow MWD (Mw/Mn = 1.20) (Table 4, no. 10 and Supporting Information, Figure S26). As expected, the chain extensions of PMA and PDMA were successful with MA, DMA, and St to yield well-defined PMA-bPDMA, PMA-b-PSt, PDMA-b-PMA, and PDMA-b-PSt block copolymers, respectively (Table 4, nos. 12, 13, 15, and 16, and Supporting Information, Figures S27−30). Finally, the synthesis of PNVP-b-PVAc was achieved using catalyst concentration of 10 ppm relative to monomer (Table 4, no. 18 and Supporting Information, Figure S31). 6. Synthesis of Multiblock Copolymers via Photoinduced Living Polymerization. To further investigate the robustness of the catalyst, successive chain extensions of PMA were performed to generate a decablock P(MA)10 copolymer without supplementary addition of catalyst. We first synthesized a PMA macroinitiator (Mn,GPC = 8560 g/mol) by polymerization of MA in the presence of BTPA and 5 ppm of catalyst for 2 h in DMSO. NMR confirmed full monomer conversion (>98%) in the first step. For the second block, MA in a degassed DMSO solution was then added under nitrogen to the macroinitiator, and the polymerization was allowed to continue for a further 4 h to reach full monomer conversion (Supporting Information, Table S5). This process was repeated several times until the formation of the high-order multiblock polymers with high molecular weight (Mn,GPC ≈ 82 000 g/mol, Mw/Mn ≈ 1.40) was obtained (Figure 5). To our knowledge, this is the first time that such high molecular weight block copolymers were obtained using an iterative approach under light. In previous studies, short block polymers, with a typical Mn,block ranging from 500 to 5000 g/mol, were achieved using conventional C/LRP (copper(0)-mediated polymerization and RAFT polymerization).67,68 GPC analysis of the MWDs confirmed successful chain extensions as manifested by clear shifts to higher molecular weights in each step. In addition, the MWD remained narrow (Mw/Mn ≈ 1.40) even after 10 chain extensions (Figure 5B). Mn,GPC’s were in good agreement with the theoretical values, although the formation of some low molecular weight tailings was observed after 4−5 cycles (Figure 5A). Semiquantitative analysis using GPC traces was also employed to estimate the proportion of dead polymers after each chain extension using a previously published procedure.68,69 The method is based on the hypothesis that the low molecular weight tails are attributed to the dead polymers. To estimate the living chains by using this method, we converted the molecular weight distributions to the corresponding number distributions (Supporting Information, Figure S32). Figure 5C showed the plot of livingness versus number of chain extensions. Unfortunately, we observed a gradual decrease of the livingness after each cycle. Nevertheless, the livingness (i.e., number of living polymer chains) was greater than 65% after 10 chain extensions, which represent almost 90 wt % of living polymer. This is a remarkable result considering the lack of optimizations to the system, that is, concentration of catalyst, light intensity, and polymerization time. In conventional controlled radical polymerization techniques, the livingness is usually less than 90% for each chain extension.68−70 These experimental results demonstrate the extreme robustness and efficiency of the

Figure 8. Preparation of block copolymers in the presence of oxygen. (A) The “ON”/“OFF” experiment for preparing triblock copolymer PMA-b-PtBuA-b-PnBuA; (B) molecular weight distributions of triblock copolymer PMA-b-PtBuA-b-PnBuA; and (C) molecular weight distribution of diblock of PMMA-b-PMMA.

macroinitiator was prepared using CPADB and chain extended in the presence of St, MA, and HPMA using a concentration of photoredox catalyst of 5 ppm relative to the monomer. Surprisingly, the chain extensions of PMMA with St and MA were unsuccessful, as NMR and GPC did not reveal monomer conversion or a shift of MWD, respectively. This is a major difference between our process and conventional RAFT or ATRP polymerization techniques. Indeed, PMMA can be easily chain extended with MA and St using the conventional RAFT polymerization process. One possible explanation is that the photoredox catalyst cannot activate PMA−S(CS)−Ph endgroup. Nevertheless, the chain extension of PMMA with HPMA was successful via photoinduced living polymerization, resulting in the synthesis of PMMA-b-HPMA with a narrow MWD (Mw/Mn < 1.2) (Table 4, no. 4 and Supporting Information, Figure S24). PHPMA was also successfully chain extended in the presence of MMA to yield PHPMA-b-PMMA diblock copolymer (Table 4, no. 7 and Supporting Information, Figure S25). The chain extension of PSt macroinitiator with MMA was uncontrolled, resulting in a much higher molecular weight polymer than the theoretical values with a broad MWD (Table 5516

dx.doi.org/10.1021/ja501745g | J. Am. Chem. Soc. 2014, 136, 5508−5519

Journal of the American Chemical Society

Article

exposure time (Figure 7A and C). Interestingly, the slopes of ln([M]0/[M]t) (apparent propagation constant, kpapp) versus time for the polymerizations in the presence of air were almost the same as the degassed reactions after the inhibition period (Figure 7A and C), which indicated that both photocatalyst and thiocarbonylthio compounds were not degraded during the oxygen reduction period. In addition, the evolutions of Mn,GPC values versus exposure time were in good agreement with those in the absence of oxygen (degassed system) and theoretical values (Figure 7B and D). GPC showed a shift of the molecular weight distribution to higher molecular weight with a narrow polydispersity (Supporting Information, Figure S33). To further investigate the livingness (i.e., the end-group fidelity) and the robustness of the catalyst in a nondegassed environment, successive chain extensions of PMA and PMMA were performed to generate a diblock of PMMA-b-PMMA and a triblock of PMA-b-PtBuA-b-PnBuA copolymers without degassing the solutions (Figure 8). In this approach, we decided to use an iterative process as described in the previous paragraph. To our knowledge, it is the first time that block copolymers were obtained without purification and also degassing between each chain extension. For each chain extension, we added a nondegassed solution containing monomer and solvent. Subsequently, the solution was placed under a 4.8 W blue LED light for 7 h (MA) and 24 h (MMA) to obtain high monomer conversion (>98% determined by 1H NMR). The solutions then were placed in the dark to avoid the formation of dead polymers during monomer conversion analysis. After confirmation of full monomer conversion, a new aliquot of monomer and solvent was added to the mixture. After an inhibition period of 2−3 h, the polymerization proceeded until full monomer conversion. GPC revealed the formation of welldefined block copolymers with a narrow MWD (Mw/Mn < 1.10). Purified copolymers were finally analyzed by NMR to determine the exact composition (Supporting Information, Figure S34).

catalyst in this photoinduced living polymerization to make multiple chain extensions. 7. Polymerization in the Presence of Air. Another unique property of this process is the possibility to perform the polymerization without degassing the reaction mixtures. It is well-known that oxygen is detrimental to radical polymerizations, as oxygen is an excellent radical scavenger. Conventional free radical and controlled/living radical polymerization techniques, including ATRP,49,50 RAFT,57 and NMP,70 are susceptible to trace amounts of oxygen and require deoxygenation procedures, such as degassing with nitrogen or several freeze−pump−thaw cycles (except for ARGET,71−73 SET-LRP, 64,74,75 and SARA process;76 however, these techniques require specific conditions and can only polymerize a narrow range of monomer families (i.e., usually acrylate)). Construction of an oxygen-free environment could be challenging for specific industrial applications, such as surface modifications, miniemulsion polymerization, coatings, etc. We hypothesized that the excellent reductive qualities of fac[Ir(ppy)3] catalyst would be able to overcome the lack of deoxygenation by reducing oxygen into inactive species. As the redox potential of O2 in acetonitrile in the presence of water is around +1.23 V,77,78 we hypothesized that the reduction of oxygen into hydroxide ion (OH) was possible in our system. In a closed and sealed vessel, the oxygen would slowly be consumed and eliminated, while a small amount of hydroxide ion will be produced as a side product. In our first attempts, we decided to test our hypothesis for the polymerization of MMA and MA. The polymerizations of MMA and MA were performed in a sealed but nondegassed vessel of 4 mL using a total liquid volume of 3 mL (50/50 (v/ v) of solvent/monomer) under a 4.8 W blue LED light. After 24 h, the reaction solutions were analyzed by 1H NMR and GPC. NMR revealed a monomer conversion of 99% and 50% for MA and MMA, respectively, while GPC showed the presence of PMA and PMMA with very good control of the molecular weight in good agreement with the theoretical values and Mw/Mn (